I'm sure someone smarter than me has a toolkit for these things, I just don't know where to find it.
Store-and-forward-wise, NNCP is designed for this, but it's not widespread yet.
It takes me about a month after a reinstall or new machine, to feel like I've really spread my wings and have everything installed that I initially forgot about. So I guess the recommendation would be "daily-drive it for a month before refrigerating it". And at that point, you might as well just make it your everyday machine.
I had no idea. I guess this refers to app-store rules? Related to https://developer.android.com/guide/topics/manifest/applicat... and https://developer.apple.com/documentation/security/preventin... ?
Or is it even hard to browse to http webpages? (No problem on iOS that I see.)
For anything that takes even those out (eg. a "Big One" quake in California), you fall back to radio hams and autonomous radio links for the disaster services.
Real time chat: wild unsecure simplicity proven to run anywhere (IRC), bells & whistles with contemporary security (Matrix), some mesh native that almost no one knows ? What about post-disaster onboarding of actual users ?
Store & forward messaging: SMTP & friends may work nicely, but with actually distributed servers - in each local disaster POP. Also needs timeout and retry parameters to keeping stuff in queues practically forever.
Forums: anything better than ol' NNTP ? Other protocols merely adopted intermittent indirect connectivity - NNTP was born in it !
Is anything more sophisticated or more interactive realistic for actual disaster ?
An onboarding kit with clients for each major OS (à la AOL CDROM !) might be handy too, for snearkernet distribution over USB dongles.